Use subplots, fix #230

[Robinlovelace]

Update isn't perfect but it works:

There may be other examples where we need to do this so putting as PR, feel free to merge @michaeldorman

[michaeldorman]

Thanks for looking into this @Robinlovelace ! It's a great idea to make the figures as single image files, and solves the issue, but it also makes the code significantly more complicated with no added value to the readers. From https://github.com/geocompx/geocompy/issues/230#issuecomment-2156001430 it seems like this is going to be fixed in quarto. The question is whether we should

• rely on the future fix and keep using the multi-column layout, or
• switch to your solution which would make the code longer but more independent of quarto

I suggest the former, unless this is not fixed in time and then we'll have no choice but to use the latter

[Nowosad]

@michaeldorman there is also a third possible solution (I do not know if the best): you could show the shorter code, but instead of running it -- you could just provide an image in that place (just remember to redo the code when the quarto fix is ready).

[michaeldorman]

@michaeldorman there is also a third possible solution (I do not know if the best): you could show the shorter code, but instead of running it -- you could just provide an image in that place (just remember to redo the code when the quarto fix is ready).

Right, didn't think of that! This is an interesting alternative. IMHO @Robinlovelace 's solution is preferable, because the entire book shows the code as-is, so we should try to stick to that. It it's just temporary then you're right it would be nice to have this kind of setup with alternative images, but it comes at a cost of lots of work to prepare and manage the images and keep track of their reproducibility, so I'd still lean towards @Robinlovelace 's solution to keep things straightforward

[Robinlovelace]

Closing this PR because it will soon be a non-issue and fixed upstream: https://github.com/quarto-dev/quarto-cli/issues/9927#event-13105363395

[Robinlovelace]

Related thought: should we start an impact.csv for this book?